Embodied RAG++: Retrieval-Augmented Generation From What Exists to What Works in Embodied Scenes
cs.RO
Submitted: 2026-11-05
Updated: 2026-11-05
Terminology
Sources
- 3D Scene Graph: A Structure for Unified Semantics, 3D Space, and Camera
- Hydra: A Real-time Spatial Perception System for 3D Scene Graph Construction and Optimization
- ConceptGraphs: Open-Vocabulary 3D Scene Graphs for Perception and Planning
- SayPlan: Grounding Large Language Models using 3D Scene Graphs for Scalable Robot Task Planning
- SQA3D: Situated Question Answering in 3D Scenes
- Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
- From Local to Global: A Graph RAG Approach to Query-Focused Summarization
- RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval
- HippoRAG: Neurobiologically Inspired Long-Term Memory for Large Language Models
- Embodied-RAG: General Non-parametric Embodied Memory for Retrieval and Generation
- RAG-3DSG: Enhancing 3D Scene Graphs with Re-Shot Guided Retrieval-Augmented Generation
- LEANN: A Low-Storage Vector Index
- Unsupervised Dense Information Retrieval with Contrastive Learning
- Qwen2.5 Technical Report
- The Llama 3 Herd of Models
- LightRAG: Simple and Fast Retrieval-Augmented Generation
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation